An assessment on the impact of holidays on Nigeria Stock Exchange price returns using EGARCH and Prophet models under two different distributions of innovations

Abstract

In this study, the aim is to investigate the forecast performance of EGARCH and Prophet models on holidays effect of Nigeria Stock Exchange (NSE) price returns. The holidays effect are incorporated into EGARCH model to investigate whether the volatility of the series returns will decrease or increase thereby increasing the accuracy of results than using ordinary EGARCH model. On the other hand, Prophet model also called Facebook Prophet model was used to investigate whether including functions such as trend, seasonality, and holidays could accurately forecast the return series. Based on the evaluation criteria, EGARCH (under skew student’s t distribution of innovation) performed better than the Facebook Prophet model as well as EGARCH (with normality assumptions of innovations) model. Furthermore, the results revealed a positive return series for holidays that falls on Thursdays.